JalView

Optimal pairwise alignment | Algorithms

Task 1

According the task of pr 11 I used two homologous protein sequences from pr 10 and campared their pairwise global and local alignments. I used "needle" (EMBOSS) for global alignment and "water" (EMBOSS) for local alignment. Then I used "infoalign" (EMBOSS) to find out the characteristics of these alignments (see tables 1, 2). To represent the results I used Jalview (program Tcoffee with Defaults, colored by ClustalX, with Identity Threshold = 100%).

Table 1.The parametres of grobal alignment.

Name	SeqLen	AlignLen	GapLen	% of GapLen	Absolutely Ident	% of Abs. Ident	Functional Ident	% of Func. Ident
DNAK_HALWD	641	692	51	7,37	140	20,23	270	39,02
HSP7F_ENCCU	658	769	111	14,43	140	18,21	270	39,02

! To see the whole picture press on it.
Picture 1.The representation of pairwise global alignment.

Table 2.The parametres of local alignment.

Name	SeqLen	AlignLen	GapLen	% of GapLen	Absolutely Ident	% of Abs. Ident	Functional Ident	% of Func. Ident
DNAK_HALWD	581	618	37	5,99	135	21,84	254	41,10
HSP7F_ENCCU	534	618	84	13,59	135	21,84	254	41,10

! To see the whole picture press on it.
Picture 2.The representation of pairwise local alignment.

By command "needle -help -verbose" and "water -help -verbose" I found out a system of penalties which is used to make alignments (see table 3). This information is included in units "Stundart qualifiers" and "Additional qualifiers".

Table 3.Penalties for gaps in different cases.

Program	The gap open penalty inside the sequense	The gap extension penalty is added to the standard gap penalty for each residue in the gap	The terminal gap open penalty	The gap extension penalty is added to the standard terminal gap penalty for each residue in the gap
Needle	10.0 [1.0 - 100.0]	0.5 [0.0 - 10.0]	10.0 [1.0 - 100.0]	0.5 [0.0 - 10.0]
Water	10.0 [1.0 - 100.0]	0.5 [0.0 - 10.0]	NO	NO

*Here can be some special cases when you should use very low open gap penalty: "You can get this result by setting the gap open penalty to zero (or very low) and using the gap extension penalty to control gap scoring".

The comparision of pairwise global and local alignments:

• The global alignment is longer than the local one.
• The local alignment starts from the fourth position relative to the global alignment. Because of this, the numbering is shifted if you compare the alignments.
  Then there are a big block the same in both alignments (276 positions).
• After 276 position in the local alignment there is a 'block' without gaps, but at the same position (280) in the global alignment there is least 'block' and part of second sequence with gaps.
• From this point two alignments are very different. The local one includes more matching residues than the global one, and the global one includes much more parts with gaps.

Task 2

According the task I compared pairwise local alignments of two homologous and five pairs of nonhomologous proteins. The homologous protein sequences were taken from task 1. As nonhomologous protein sequences I used "my protein" (pr 1) and five proteins randomly chosen between "my classmate's proteins". You can see some infirmation about these proteins in table 4. Line with "my protein" is in bold.

Table 4. Some information about proteins which I used.

Entry name	SeqLen	Protein name	Organism
DNAK_HALWD	641	Chaperone protein DnaK	Haloquadratum walsbyi
HSP7F_ENCCU	658	Heat shock protein homolog SSE1	Encephalitozoon cuniculi
A0A0U3W9X5_9BACI	322	Fructose-1,6-bisphosphatase	Lentibacillus amyloliquefaciens
A0A0U3QLP1_9MICC	242	ADP-ribose pyrophosphatase	Arthrobacter sp.
A0A0X8CMX2_9BRAD	772	Uncharacterized protein	Bradyrhizobium sp.
A0A0U3QKR4_9MICC	444	Alpha-L-fucosidase	Arthrobacter sp.
A0A0F6CLF2_MYCGL	1269	CRISPR-associated endonuclease	Mycoplasma gallisepticum
A0A0H4VDW6_9SPHN	768	Cell division cycle protein	Erythrobacter atlanticus

Table 5.The parametres of local alignment of two homologous proteins and five pairs of nonhomologous proteins.

Name	SeqLen	AlignLen	% of SeqLen in local alignment in compare with protein length	GapLen	% of GapLen	Absolutely Ident	% of Abs. Ident	Functional Ident	% of Func. Ident
DNAK_HALWD	581	618	90,64 (581/641)	37	5,99	135	21,84	254	41,10
HSP7F_ENCCU	534	618	81,16(534/658)	84	13,59	135	21,84	254	41,10
A0A0U3W9X5_9BACI	74	76	22,98 (74/322)	2	2,63	21	27,63	31	40,79
A0A0U3QLP1_9MICC	73	76	30,17 (73/242)	3	3,95	21	27,63	31	40,79
A0A0U3W9X5_9BACI	176	223	54,66 (176/322)	47	21,08	44	19,73	75	33,63
A0A0X8CMX2_9BRAD	198	223	25,65 (198/772)	25	11,21	44	19,73	75	33,63
A0A0U3W9X5_9BACI	174	211	54,04 (174/322)	37	17,54	41	19,43	65	30,81
A0A0U3QKR4_9MICC	153	211	34,46 (153/444)	58	27,49	41	19,43	65	30,81
A0A0U3W9X5_9BACI	34	57	10,56 (34/322)	23	40,35	14	24,56	20	35,09
A0A0F6CLF2_MYCGL	57	57	4,49 (57/1269)	0	0	14	24,56	20	35,09
A0A0U3W9X5_9BACI	212	320	65,84 (212/322)	108	33,75	67	20,94	105	32,81
A0A0H4VDW6_9SPHN	315	320	41,02 (315/768)	5	1,56	67	20,94	105	32,81

The comparision of pairwise local alignments of homologous and nonhomologous sequences:

• In the local alignment of homologous sequences the percentage of amino acids of the original sequence included in the alignment is near 80%. In the local alignment of nonhomologous sequences it is most likely in the range from 5 to 40. I think this is a very important characteristic.
• The average percentage of gaps is more in alignment of homologous sequences (near 10%) than nonhomologous (more than 15% or 25 %).
• The percentage of absolute conserved in all cases are almost identical (near 20-25 %). It is a bit strange, I think there must be more absolutely ident positions in the alignment of homologous sequences.
• The percentage of functional conserved is more in alignment of homologous sequences.

Picture 3.The representation of pairwise local alignment of two nonhomologous protein sequences.

Task 3

The comparision of alignments of the cut out piece of multiple alignments and the global alignment, the cut out piece of multiple alignments and the local alignment, the global alignmetn and the local alignment:

1. Two alignments are the same: cut out piece of multiple alignments with the local alignment and the global alignmetn with the local alignment.
2. There are some difference between these alignments and cut out piece of multiple alignments with the global alignment:
   • Position 297 global - Q is mapped to gap, and local - Q is mapped to K.
   • Position 303 global - G is mapped to gap, local - column of G.
   • Position 305 global - D is mapped to K, and local - D is mapped to gap.
   • Position 310 global - D is mapped to N, and local - D is mapped to gap.
   • So in global alignment the piece "KETEFNG" is moved from 297 to 305 relative to the local alignment.

! To see the whole picture press on it.
Picture 4.The representation of alignment: the cut out piece of multiple alignments and the global alignment.

Picture 5.The representation of alignment: the cut out piece of multiple alignments and the local alignment.

Picture 6.The representation of alignment: the global alignmetn and the local alignment.

The conclusion:
I think that the local alignment is more reliable than global and the cut out piece of multiple alignment. Reasons: 1. the site which was discussed above contains more conserved columns in local alignment than in global; 2. this site contains two starts of gaps in the global alignment and only one in the local one.